How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
hypnguyen1209CLAUDE.mdhypnguyen1209/offensive-claude/CLAUDE.md
paths and engagement strategies | | exploit-researcher | CVE research and exploitation chain development | | security-reviewer | Deep code security audit | | reverse-engineer | Binary analysis and vulnerability discovery | | ai-researcher | AI/ML architecture, training
smart-mcp-proxyCLAUDE.mdsmart-mcp-proxy/mcpproxy-go/CLAUDE.md
variant; Spec 085 pre-dispatch arg validation with self-healing `invalid_params` errors) · `code_execution` (sandboxed JS, on by default since v0.66.0) · `upstream_servers` (CRUD, Spec 049) · `quarantine_security` (Spec
superbasicstudioCLAUDE.mdsuperbasicstudio/claude-conductor/CLAUDE.md
move & summarize ## Project Architecture This is a **documentation framework** designed to work with Claude Code AI assistant. The system provides: ### Core Component - **CONDUCTOR.md** (`CONDUCTOR.md:1-437`) - Master template that defines
superbasicstudioCLAUDE.mdsuperbasicstudio/claude-conductor/templates/CLAUDE.md
this repository. ## Critical Context (Read First) - **Tech Stack**: [List core technologies] - **Main File**: [Primary code file and line count] - **Core Mechanic**: [One-line description] - **Key Integration**: [Important external services] - **Platform
MeridionaCLAUDE.mdMeridiona/meridian/CLAUDE.md
code — other agents may be working on the codebase in parallel - NEVER merge a PR automatically — open/update PRs as needed, but leave the actual merge to a human reviewer - NEVER
openaccountantsCLAUDE.mdopenaccountants/openaccountants/CLAUDE.md
those sections with the exact marker `**AUDIT FLASH POINT**`. CPAs scan for this when reviewing. Examples: §174 R&D capitalization, §1402(a)(13) limited-partner SE (post-Soroban), §183 hobby
ai-sdlc-frameworkCLAUDE.mdai-sdlc-framework/ai-sdlc/CLAUDE.md
there for operator routing — do not self-authorize. ### Reviewer gate (AISDLC-308) The `code-reviewer` and `test-reviewer` subagents check for scope-creep candidates in every
VasiHemanthCLAUDE.mdVasiHemanth/tokentelemetry/.claude/CLAUDE.md
session link under the "Generated with Claude Code" line. The `🤖 Generated with [Claude Code]` attribution itself is fine and should stay. It's only the session URL that must
AnthropicCLAUDE.mdanthropics/anthropic-sdk-ruby/CLAUDE.md
lookup table), not to restate what the code does or narrate its history — over-documenting the obvious is its own review finding. Param docs read "Defaults to `ENV["…"]`, otherwise
zamesinCLAUDE.mdzamesin/Next-Move-Theory-Canon-and-Skills/CLAUDE.md
repair the step that stops a new segment from reaching value (Stripe's no-code drop-in for the non-engineer). - **Lower one of the six costs** — and find which
db-lyonCLAUDE.mddb-lyon/ue-mcp/CLAUDE.md
CLAUDE.md Operating guide for Claude Code (and any AI agent) working in this repo. Shared team knowledge - checked into the repo so every contributor gets the same rulebook. ## Repo
liangdabiaoCLAUDE.mdliangdabiao/Claude-Code-Stock-Deep-Research-Agent/CLAUDE.md
this repository. ## Project Overview This repository contains **two specialized deep research frameworks** for Claude Code: 1. **Stock Investment Research** (股票投资尽调) - A specialized 8-phase investment due diligence framework for analyzing
cbcoutinhoCLAUDE.mdcbcoutinho/nextcloud-mcp-server/CLAUDE.md
list exactly what Sonar flagged on the PR's new code and read the gate — do this *inside the review loop*, before merge, and fix the findings: ```bash sonar list
timwuhaotianCLAUDE.mdtimwuhaotian/the-pair/CLAUDE.md
agents working together on a coding task: - **Mentor** — read-only planner/reviewer - **Executor** — code writer / command runner The Rust backend spawns local provider CLIs (`opencode`, `claude`, `codex`, `gemini`, `kimi`, `pi`, `kiro
mezmoCLAUDE.mdmezmo/aura/CLAUDE.md
flags), and the architecture overview. - `CONTRIBUTING.md` - Contribution process only: CLA, workflow, code quality standards, commit conventions, PR and review process. Build/test details belong in `DEVELOPMENT.md`; link, don't duplicate. - `docs
milady-aiCLAUDE.mdmilady-ai/milady/CLAUDE.md
meaningful redesign. **Hard rules:** - Every page has a screenshot AND a `manual-review/ .md` checked in. - `needs-work` / `broken` pages block "done." - Brand orange is accent only. No blue anywhere
spencermarxCLAUDE.mdspencermarx/open-code-review/CLAUDE.md
lint` — an accidental `dashboard → cli` (or `shared → app`) import fails the build, not just review. Code both apps need (persistence, domain/state, config, cross-platform utilities) lives in dedicated library packages
0xquintoCLAUDE.md0xquinto/bcherny-claude/CLAUDE.md
lint # Lint all files npm run format # Format code # Git workflow git status # Check current state git diff # Review changes before commit ``` ## Self-Improvement After every correction or mistake, update
coffeegrind123CLAUDE.mdcoffeegrind123/gemini-for-claude-code/CLAUDE.md
successful software development tasks. ## Task Management For complex or multi-step tasks, Claude Code will use: * **TodoWrite**: To create a structured task list, breaking down the work into manageable steps
BelkinsCLAUDE.mdBelkins/ai-dive-deep/CLAUDE.md
Working agreement Most code here is written by AI and reviewed by humans. Optimise every change for reviewability. - **Never commit to the main branch.** Branch first: ` / - `, e.g. `feat/31-leads-import`; omit
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
Build and test commands, the project's layout, conventions that aren't obvious from the code, and mistakes to avoid. Short files tend to work better than long ones.
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.
